wiki:WeeklyMinutes20120605

BIND 10 Team Call

2012-06-05

ACTIONS

Attendees

  • Aharen
  • Jinmei
  • Jeremy
  • Kevin
  • Kambe
  • Larissa
  • Jelte
  • Mukund

Note

The meeting was held kind of ad-hoc, without a real agenda, so the topics below were mostly thought of on the fly.

Topics

Logging notices

Important information but not a warning. Some may not log INFO as too noisy, so what level to use? Maybe we can introduce a NOTICE level for such messages. See http://bind10.isc.org/ticket/1890 for an example (configuration file was moved). Currently: DEBUG, INFO, WARN, ERROR, FATAL, NONE. syslog(7) has: LOG_NOTICE Conditions that are not error conditions, but should possibly be handled specially. (It is between WARNING and INFO.) No NOTICE defined in log4cplus, but does have available "numbers".

jinmei: suspect will always have this type of issue that level is not perfect.

(Side notes: syslog has even more levels. log4cplus does not pre-define this NOTICE level, but the predefined levels there are 10000 integers apart, plenty of space to add your own.)

Jelte: I generally find it quite hard to pick a level when i'm adding a message, and the question here should be 'would adding a NOTICE level make the choice easier or harder (and would it result in more or less discussions about loglevels). After the explanation about 1890 i think we should probably change the message there to WARN, not INFO, if we do not add NOTICE.

Michal: I think that we have too many already, and I often find it hard to choose between them. For me two sounds nice :)

The general consensus seems to be that we do not really need it at the moment. The plan is to keep this discussion in our heads, and if we find more messages where it is hard to choose between INFO and WARN, we will reconsider adding it.

There is a plan to clean up the levels and messages before the september release, that would be a good time to decide on this.

jinmei: should we add a wiki page about controversial log messages?

General agreement.

Planning

Michal: should we start research on resolver?

jinmei: yes, shane had planned on sending an email about this

larissa: we do need to start it, but not in this sprint. The plan is also to get a bit more resources for that

Jelte: if it's in next sprint i'll want to know this week (for sprint planning)

Consensus:

We should try to remind shane about this

Not necessarily needed this week but we need to not forget about it completely.

DDNS development status

Jelte: I was looking at the sprint progress earlier today, and we seem to be doing pretty well; about as much tickets left new as we have closed now (including the added ones). A few more new ones have been added, but those seem not critical. In terms of the DDNS feature specifically, it looks as we are pretty well covered; there are known issues to be resolved but the core work is done and the biggest issues are being adressed now. So after this sprint it should be pretty much working, a bit rough around the edges perhaps, and maybe not completely feature-complete (for instance TCP support), but certainly workable.

AOB

No OB

Over at 15:01 UTC.

Last modified 5 years ago Last modified on Jun 5, 2012, 3:22:49 PM